hide long namesshow long names
hide short namesshow short names
Integer type:  int32  int64  nag_int  show int32  show int32  show int64  show int64  show nag_int  show nag_int

PDF version (NAG web site, 64-bit version, 64-bit version)
Chapter Contents
Chapter Introduction
NAG Toolbox

NAG Toolbox: nag_lapack_zgtrfs (f07cv)

Purpose

nag_lapack_zgtrfs (f07cv) computes error bounds and refines the solution to a complex system of linear equations AX = B AX=B  or ATX = B ATX=B  or AHX = B AHX=B , where A A  is an n n  by n n  tridiagonal matrix and X X  and B B  are n n  by r r  matrices, using the LU LU  factorization returned by nag_lapack_zgttrf (f07cr) and an initial solution returned by nag_lapack_zgttrs (f07cs). Iterative refinement is used to reduce the backward error as much as possible.

Syntax

[x, ferr, berr, info] = f07cv(trans, dl, d, du, dlf, df, duf, du2, ipiv, b, x, 'n', n, 'nrhs_p', nrhs_p)
[x, ferr, berr, info] = nag_lapack_zgtrfs(trans, dl, d, du, dlf, df, duf, du2, ipiv, b, x, 'n', n, 'nrhs_p', nrhs_p)

Description

nag_lapack_zgtrfs (f07cv) should normally be preceded by calls to nag_lapack_zgttrf (f07cr) and nag_lapack_zgttrs (f07cs). nag_lapack_zgttrf (f07cr) uses Gaussian elimination with partial pivoting and row interchanges to factorize the matrix A A  as
A = PLU ,
A=PLU ,
where P P  is a permutation matrix, L L  is unit lower triangular with at most one nonzero subdiagonal element in each column, and U U  is an upper triangular band matrix, with two superdiagonals. nag_lapack_zgttrs (f07cs) then utilizes the factorization to compute a solution, X^ , to the required equations. Letting x^  denote a column of X^ , nag_lapack_zgtrfs (f07cv) computes a component-wise backward error, β β , the smallest relative perturbation in each element of A A  and b b  such that x^  is the exact solution of a perturbed system
(A + E) = b + f , with  |eij| β |aij| , and  |fj| β |bj| .
(A+E) x^=b+f , with  |eij| β |aij| , and  |fj| β |bj| .
The function also estimates a bound for the component-wise forward error in the computed solution defined by max |xixi^| / max |xi^| max | xi - xi^ | / max |xi^| , where x x  is the corresponding column of the exact solution, X X .

References

Anderson E, Bai Z, Bischof C, Blackford S, Demmel J, Dongarra J J, Du Croz J J, Greenbaum A, Hammarling S, McKenney A and Sorensen D (1999) LAPACK Users' Guide (3rd Edition) SIAM, Philadelphia http://www.netlib.org/lapack/lug

Parameters

Compulsory Input Parameters

1:     trans – string (length ≥ 1)
Specifies the equations to be solved as follows:
trans = 'N'trans='N'
Solve AX = BAX=B for XX.
trans = 'T'trans='T'
Solve ATX = BATX=B for XX.
trans = 'C'trans='C'
Solve AHX = BAHX=B for XX.
Constraint: trans = 'N'trans='N', 'T''T' or 'C''C'.
2:     dl( : :) – complex array
Note: the dimension of the array dl must be at least max (1,n1)max(1,n-1).
Must contain the (n1)(n-1) subdiagonal elements of the matrix AA.
3:     d( : :) – complex array
Note: the dimension of the array d must be at least max (1,n)max(1,n).
Must contain the nn diagonal elements of the matrix AA.
4:     du( : :) – complex array
Note: the dimension of the array du must be at least max (1,n1)max(1,n-1).
Must contain the (n1)(n-1) superdiagonal elements of the matrix AA.
5:     dlf( : :) – complex array
Note: the dimension of the array dlf must be at least max (1,n1)max(1,n-1).
Must contain the (n1)(n-1) multipliers that define the matrix LL of the LULU factorization of AA.
6:     df( : :) – complex array
Note: the dimension of the array df must be at least max (1,n)max(1,n).
Must contain the nn diagonal elements of the upper triangular matrix UU from the LULU factorization of AA.
7:     duf( : :) – complex array
Note: the dimension of the array duf must be at least max (1,n1)max(1,n-1).
Must contain the (n1)(n-1) elements of the first superdiagonal of UU.
8:     du2( : :) – complex array
Note: the dimension of the array du2 must be at least max (1,n2)max(1,n-2).
Must contain the (n2)(n-2) elements of the second superdiagonal of UU.
9:     ipiv( : :) – int64int32nag_int array
Note: the dimension of the array ipiv must be at least max (1,n)max(1,n).
Must contain the nn pivot indices that define the permutation matrix PP. At the iith step, row ii of the matrix was interchanged with row ipiv(i)ipivi, and ipiv(i)ipivi must always be either ii or (i + 1)(i+1), ipiv(i) = iipivi=i indicating that a row interchange was not performed.
10:   b(ldb, : :) – complex array
The first dimension of the array b must be at least max (1,n)max(1,n)
The second dimension of the array must be at least max (1,nrhs)max(1,nrhs)
The nn by rr matrix of right-hand sides BB.
11:   x(ldx, : :) – complex array
The first dimension of the array x must be at least max (1,n)max(1,n)
The second dimension of the array must be at least max (1,nrhs)max(1,nrhs)
The nn by rr initial solution matrix XX.

Optional Input Parameters

1:     n – int64int32nag_int scalar
Default: The first dimension of the arrays b, x The dimension of the arrays d, df, ipiv.
nn, the order of the matrix AA.
Constraint: n0n0.
2:     nrhs_p – int64int32nag_int scalar
Default: The second dimension of the arrays b, x.
rr, the number of right-hand sides, i.e., the number of columns of the matrix BB.
Constraint: nrhs0nrhs0.

Input Parameters Omitted from the MATLAB Interface

ldb ldx work rwork

Output Parameters

1:     x(ldx, : :) – complex array
The first dimension of the array x will be max (1,n)max(1,n)
The second dimension of the array will be max (1,nrhs)max(1,nrhs)
ldxmax (1,n)ldxmax(1,n).
The nn by rr refined solution matrix XX.
2:     ferr(nrhs_p) – double array
Estimate of the forward error bound for each computed solution vector, such that jxj / jferr(j)x^j-xj/x^jferrj, where jx^j is the jjth column of the computed solution returned in the array x and xjxj is the corresponding column of the exact solution XX. The estimate is almost always a slight overestimate of the true error.
3:     berr(nrhs_p) – double array
Estimate of the component-wise relative backward error of each computed solution vector jx^j (i.e., the smallest relative change in any element of AA or BB that makes jx^j an exact solution).
4:     info – int64int32nag_int scalar
info = 0info=0 unless the function detects an error (see Section [Error Indicators and Warnings]).

Error Indicators and Warnings

  info = iinfo=-i
If info = iinfo=-i, parameter ii had an illegal value on entry. The parameters are numbered as follows:
1: trans, 2: n, 3: nrhs_p, 4: dl, 5: d, 6: du, 7: dlf, 8: df, 9: duf, 10: du2, 11: ipiv, 12: b, 13: ldb, 14: x, 15: ldx, 16: ferr, 17: berr, 18: work, 19: rwork, 20: info.
It is possible that info refers to a parameter that is omitted from the MATLAB interface. This usually indicates that an error in one of the other input parameters has caused an incorrect value to be inferred.

Accuracy

The computed solution for a single right-hand side, x^ , satisfies an equation of the form
(A + E) = b ,
(A+E) x^=b ,
where
E = O(ε)A
E=O(ε)A
and ε ε  is the machine precision. An approximate error bound for the computed solution is given by
(x)/(x) κ(A) (E)/(A) ,
x^-x x κ(A) E A ,
where κ(A) = A1 A κ(A)=A-1 A , the condition number of A A  with respect to the solution of the linear equations. See Section 4.4 of Anderson et al. (1999) for further details.
Function nag_lapack_zgtcon (f07cu) can be used to estimate the condition number of A A .

Further Comments

The total number of floating point operations required to solve the equations AX = B AX=B  or ATX = B ATX=B  or AHX = B AHX=B  is proportional to nr nr . At most five steps of iterative refinement are performed, but usually only one or two steps are required.
The real analogue of this function is nag_lapack_dgtrfs (f07ch).

Example

function nag_lapack_zgtrfs_example
trans = 'No transpose';
dl = [ 1 - 2i;
      1 + 1i;
      2 - 3i;
      1 + 1i];
d = [ -1.3 + 1.3i;
      -1.3 + 1.3i;
      -1.3 + 3.3i;
      -0.3 + 4.3i;
      -3.3 + 1.3i];
du = [ 2 - 1i;
      2 + 1i;
      -1 + 1i;
      1 - 1i];
dlf = [ -0.78 - 0.26i;
      0.162 - 0.486i;
      -0.04516923076923077 - 0.0009538461538460446i;
      -0.3978553846153843 - 0.05620307692307711i];
df = [ 1 - 2i;
      1 + 1i;
      2 - 3i;
      1 + 1i;
      -1.339863692307691 + 0.2875264615384604i];
duf = [ -1.3 + 1.3i;
      -1.3 + 3.3i;
      -0.3 + 4.3i;
      -3.3 + 1.3i];
du2 = [ 2 + 1i;
      -1 + 1i;
      1 - 1i];
ipiv = [int64(2);3;4;5;5];
b = [ 2.4 - 5i,  2.7 + 6.9i;
      3.4 + 18.2i,  -6.9 - 5.3i;
      -14.7 + 9.7i,  -6 - 0.6i;
      31.9 - 7.7i,  -3.9 + 9.3i;
      -1 + 1.6i,  -3 + 12.2i];
x = [ 1 + 1i,  2 - 1i;
      3 - 1i,  1 + 2i;
      4 + 5i,  -1 + 1i;
      -1 - 2i,  2 + 1i;
      1 - 1i,  2 - 2i];
[xOut, ferr, berr, info] = nag_lapack_zgtrfs(trans, dl, d, du, dlf, df, duf, du2, ipiv, b, x)
 

xOut =

   1.0000 + 1.0000i   2.0000 - 1.0000i
   3.0000 - 1.0000i   1.0000 + 2.0000i
   4.0000 + 5.0000i  -1.0000 + 1.0000i
  -1.0000 - 2.0000i   2.0000 + 1.0000i
   1.0000 - 1.0000i   2.0000 - 2.0000i


ferr =

   1.0e-13 *

    0.5256
    0.7018


berr =

   1.0e-16 *

    0.2168
    0.5921


info =

                    0


function f07cv_example
trans = 'No transpose';
dl = [ 1 - 2i;
      1 + 1i;
      2 - 3i;
      1 + 1i];
d = [ -1.3 + 1.3i;
      -1.3 + 1.3i;
      -1.3 + 3.3i;
      -0.3 + 4.3i;
      -3.3 + 1.3i];
du = [ 2 - 1i;
      2 + 1i;
      -1 + 1i;
      1 - 1i];
dlf = [ -0.78 - 0.26i;
      0.162 - 0.486i;
      -0.04516923076923077 - 0.0009538461538460446i;
      -0.3978553846153843 - 0.05620307692307711i];
df = [ 1 - 2i;
      1 + 1i;
      2 - 3i;
      1 + 1i;
      -1.339863692307691 + 0.2875264615384604i];
duf = [ -1.3 + 1.3i;
      -1.3 + 3.3i;
      -0.3 + 4.3i;
      -3.3 + 1.3i];
du2 = [ 2 + 1i;
      -1 + 1i;
      1 - 1i];
ipiv = [int64(2);3;4;5;5];
b = [ 2.4 - 5i,  2.7 + 6.9i;
      3.4 + 18.2i,  -6.9 - 5.3i;
      -14.7 + 9.7i,  -6 - 0.6i;
      31.9 - 7.7i,  -3.9 + 9.3i;
      -1 + 1.6i,  -3 + 12.2i];
x = [ 1 + 1i,  2 - 1i;
      3 - 1i,  1 + 2i;
      4 + 5i,  -1 + 1i;
      -1 - 2i,  2 + 1i;
      1 - 1i,  2 - 2i];
[xOut, ferr, berr, info] = f07cv(trans, dl, d, du, dlf, df, duf, du2, ipiv, b, x)
 

xOut =

   1.0000 + 1.0000i   2.0000 - 1.0000i
   3.0000 - 1.0000i   1.0000 + 2.0000i
   4.0000 + 5.0000i  -1.0000 + 1.0000i
  -1.0000 - 2.0000i   2.0000 + 1.0000i
   1.0000 - 1.0000i   2.0000 - 2.0000i


ferr =

   1.0e-13 *

    0.5256
    0.7018


berr =

   1.0e-16 *

    0.2168
    0.5921


info =

                    0



PDF version (NAG web site, 64-bit version, 64-bit version)
Chapter Contents
Chapter Introduction
NAG Toolbox

© The Numerical Algorithms Group Ltd, Oxford, UK. 2009–2013